Mixtile Core 3588E

 

System-on-Module

Up to 6 TOPs AI Performance

All Pins are compatible with NVIDIA Jetson TX2 NX*

*Use NVIDIA Jetson TX2 NX-supported Carrier Boards. Others may be incompatible.

Mixtile Core 3588E system-on-module powered by Rockchip RK3588 SoC with a 260-pin-SO-DIMM connector compatible with NVIDIA Jetson TX2 NX. It is designed to support multiple applications based on ARM high-performance and low-power processors.

Benchmarks of Core 3588E

Support Large Language Model (LLM)

Boasting up to 6 TOPs AI performance, Core 3588E supports Large Language Model (LLM) such as GPT, LLaMA, ChatGLM, and Bloom. It also enables multi-chip parallel acceleration for handling very large model inferences. Users can even create an ARM Server with clusters of up to 75 Core 3588E SoMs in a standard 2U, 19-inch chassis.

Core 3588E Heatsink

Core 3588E Heatsink is a cooling fan dedicated to Mixtile Core 3588E SoM, boasting a dual-ball bearing with four-wire PWM temperature control. This fan can dynamically adjust its speed between 0 and 5000 RPM to cater to the module’s requirements. Moreover, it offers impressive longevity, with a 70,000-hour lifespan, ensuring your Core 3588E module’s stable and smooth operation, even during heavy workloads.

Industrial-grade Mixtile Core 3588JE

Consider the Mixtile Core 3588JE for robust and reliable solutions in demanding industrial environments. It has the same pinout as the Mixtile Core 3588E but uses industrial-grade components that can withstand temperatures from -40 °C to 85°C for extended periods.

Industrial-grade Mixtile Core 3588JE

Consider the Mixtile Core 3588JE for robust and reliable solutions in demanding industrial environments. It has the same pinout as the Mixtile Core 3588E but uses industrial-grade components that can withstand temperatures from -40 °C to 85°C for extended periods.

In the Press

Technical Specifications

Technical Detail
CPU Quad-core Cortex-A76 + Quad-core Cortex-A55, Neon and FPU
GPU Arm Mali-G610 MC4 with support for OpenGL ES3.2, OpenCL 2.2, Vulkan1.2
NPU RK NN, 6 TOPS NPU with support for TensorFlow, Caffe, TFLite, PyTorch, ONNX NN, Android NN
Memory 4 GB, 8 GB, 16 GB, or 32 GB 64-bit LPDDR4
Storage 32 GB, 64 GB, 128 GB, or 256 GB eMMC 5.1 flash storage
Supported OS Debian 11, Android 11, Ubuntu 22.04, and Armbian 23.07
Supply power 5 V DC
Video codec 8Kp60 video decoding (H.264/H.265/VP9)
8Kp30 video encoding (H.264/H.265)
Display 1x HDMI interface + 1x DP/eDP combo interface, up to 7680 x 4320@60 Hz for HDMI and DP, and 3840 x 2160@60 Hz for eDP
Camera 3x 4-lane or 5x 2-lane MIPI CSI interfaces @ 2.5 Gbps/lane
ISP

48M ISP, HDR

Network 10/100/1000 BASE-T
USB 1x USB 3.0 (Gen1), 3x USB 2.0
PCIe PCIe 3.0 x4 + PCIe 2.1 x1
Others – UART DEBUG x1, UART+flow control x2
– SPI x2
– I2C x4
– CAN x1
– I2S x4
– SD 4.0, SDHOST 4.0, and SDIO 3.0
– PWM x3, GPIO x15
Connector 260-pin SO-DIMM edge connector, compatible with Jetson TX2 NX
OTP Size: 8K
Optional accessory Heatsink fan
Dimensions

69.6 mm x 45 mm

Temperature – Operating: 0°C to 80°C
– Storage: -20°C to 85°C
Relative humidity – Operating: 10% to 90%
– Storage: 5% to 95%

 

Block Diagram
Ordering Options
Dram
  • 4/8/16/32 GB 64-bit LPDDR4
System Flash
  • 32/64/128/256 GB eMMC